Table 2.
Female gender and pre-existing dyspnea were associated with persistent symptomatology in patients previously hospitalized for acute COVID-19.
| Dependent Variable: Any Symptom | Dependent Variable: Dyspnea | ||||||
|---|---|---|---|---|---|---|---|
| Parameter Estimates | Variable | Adjusted OR | 95% CI |
p- Value |
Adjusted OR | 95% CI |
p- Value |
| β0 | Intercept | 7.56 | 0.65–97.36 | 0.11 | 2.32 | 0.15–35.5 | 0.5 |
| β1 | OPN (ng/μl) | 1.02 | 0.99–1.05 | 0.13 | 1.02 | 0.99–1.06 | 0.1 |
| β2 | Age(y) | 0.97 | 0.93–1.006 | 0.1 | 0.97 | 0.92–1.01 | 0.1 |
| β3 | Male gender | 0.31 | 0.12–0.77 | 0.01 | 0.39 | 0.16–0.94 | 0.03 |
| β4 | BMI ≥ 30 (kg/m2) | 1.76 | 0.84–3.77 | 0.14 | 1.43 | 0.63–3.28 | 0.4 |
| β5 | CCI > 2 | 0.97 | 0.36–2.65 | 0.95 | 1.27 | 0.41–4.12 | 0.7 |
| β6 | Non-severe acute illness | 0.87 | 0.35–2.14 | 0.76 | 0.6 | 0.21–1.65 | 0.3 |
| β7 | Pre-existing dyspnea | 10.26 | 4.07–30.07 | <0.001 | 17.64 | 7.47–46.07 | <0.001 |
Two multiple logistic regression models were developed using patient characteristics and measurements obtained during follow-up visits. Plasma OPN levels were considered as the independent variable and any symptom or dyspnea were considered as the dependent variables. Age, gender, BMI, CCI, COVID-19 severity (according to WHO) during hospitalization and the presence of dyspnea before hospitalization for COVID-19 were treated as covariates. AU-ROC curve for the first model was 0.81 (95% CI = 0.74–0.87, p < 0.001) and for the second one was 0.85 (95% CI = 0.76–0.91, p < 0.001). Estimates are presented as adjusted odds ratio (OR) and p-values <0.05 were considered significant. OPN: osteopontin, BMI: body mass index, CCI: Charlson comorbidity index, non-severe acute illness (WHO): no need for supplemental oxygen during hospitalization, WHO: World Health Organization, 95% CI: 95% confidence interval, AU-ROC curve: area under the receiver-operating characteristic curve.